KOZINSKI, Circuit Judge.
We consider whether police may conduct a search based on less than probable cause of an individual released while awaiting trial. This issue is one of first impression in our circuit. Somewhat surprisingly, it is an issue of first impression in any federal circuit and the vast majority of state courts. A lack of binding precedent does not, of course, excuse us from deciding a difficult issue when, as here, it is squarely presented.
Facts
Scott was arrested for drug possession crimes under state law and released on his own recognizance. Among the conditions of his release was consent to “random” drug testing “anytime of the day or night by any peace officer without a warrant,” and to having his home searched for drugs “by any peace officer anytime[,] day or night[,] without a warrant.”
Based on an informant’s tip, officers went to Scott’s house and administered a *890urine test. The government concedes the tip did not establish probable cause. When Scott tested positive for methamphetamine, the officers arrested him and searched his house. The search ultimately turned up a shotgun.
A federal grand jury indicted Scott for unlawfully possessing an unregistered shotgun. The district court granted Scott’s motion to suppress the shotgun and statements he had made to the officers concerning it, reasoning that the officers needed probable cause to justify the war-rantless search. The federal government took an interlocutory appeal pursuant to 18 U.S.C. § 3731.
Discussion
1. We first examine whether the searches — the drug test and the search of Scott’s house — were valid ‘because Scott consented to them as a condition of his release.
The government may detain an arrestee “to ensure his presence at trial,” Bell v. Wolfish, 441 U.S. 520, 536, 99 S.Ct. 1861, 60 L.Ed.2d 447 (1979), and may impose some conditions, such as reasonable bail, before releasing him, see United States v. Salerno, 481 U.S. 739, 754, 107 5.Ct. 2095, 95 L.Ed.2d 697 (1987). Many pre-trial detainees willingly consent to such conditions, preferring to give up some rights in order to sleep in their own beds while awaiting trial.
It may be tempting to say that such transactions — where a citizen waives certain rights in exchange for a valuable benefit the government is under no duty to grant — are always permissible and, indeed, should be encouraged as contributing to social welfare. After all, Scott’s options were only expanded when he was given the choice to waive his Fourth Amendment rights or stay in jail. Cf. Doyle v. Cont’l Ins. Co., 94 U.S. 535, 542, 24 L.Ed. 148 (1876). But our constitutional law has not adopted this philosophy wholesale. The “unconstitutional conditions” doctrine, cf. Dolan v. City of Tigard, 512 U.S. 374, 385, 114 S.Ct. 2309, 129 L.Ed.2d 304 (1994), limits the government’s ability to exact waivers of rights as a condition of benefits, even when those benefits are fully discretionary. Government is a monopoly pro-*891vider of countless services, notably law enforcement, and we live in an age when government influence and control are pervasive in many aspects of our daily lives. Giving the government free rein to grant conditional benefits creates the risk that the government will abuse its power by attaching strings strategically, striking lopsided deals and gradually eroding constitutional protections. Where a constitutional right “functions to preserve spheres of autonomy ... [unconstitutional conditions doctrine protects that [sphere] by preventing governmental end-runs around the barriers to direct commands.” Kathleen M. Sullivan, Unconstitutional Conditions, 102 Harv. L.Rev. 1413, 1492 (1989); see generally id. at 1489-1505; Richard A. Epstein, The Supreme Court, 1987 Term—Foreword: Unconstitutional Conditions, State Power, and the Limits of Consent, 102 Harv. L.Rev. 4, 21-25 (1988).
The doctrine is especially important in the Fourth Amendment context. Under modern Fourth Amendment jurisprudence, whether a search has occurred depends on whether a reasonable expectation of privacy has been violated. See Katz v. United States, 389 U.S. 347, 361, 88 S.Ct. 507, 19 L.Ed.2d 576 (1967) (Harlan, J., concurring). While the Katz principle was originally used to expand Fourth Amendment protection to cover government invasions of privacy in public places like phone booths, it can also serve to contract such protection in private places such as homes. As the Court recently explained in Kyllo v. United States, 533 U.S. 27, 121 S.Ct. 2038, 150 L.Ed.2d 94 (2001);
In assessing when a search is not a search, we have applied somewhat in reverse the principle first enunciated in Katz. ... As Justice Harlan’s oft-quoted concurrence described it, a Fourth Amendment search occurs when the government violates a subjective expectation of privacy that society recognizes as reasonable. We have subsequently *892applied this principle to hold that a Fourth Amendment search does not occur — even when the explicitly protected location of a house is concerned — unless “the individual manifested a subjective expectation of privacy in the object of the challenged search,” and “society [is] willing to recognize that expectation as reasonable.”
Id. at 32-33, 121 S.Ct. 2038 (quoting California v. Ciraolo, 476 U.S. 207, 211, 106 S.Ct. 1809, 90 L.Ed.2d 210 (1986)) (citation omitted) (alteration in original); see also United States v. Kincade, 379 F.3d 813, 873 (9th Cir.2004) (en banc) (Kozinski, J., dissenting).
The focus on subjective expectations can give rise to the following chain of logic: By assenting to warrantless house searches and random, warrantless urine tests Scott destroyed his subjective expectation of privacy, and this in turn made his searches no longer searches, depriving him of Fourth Amendment protection altogether. But the Supreme Court has resisted this logic, recognizing the slippery-slope potential of the Katz doctrine:
[I]f the Government were suddenly to announce on nationwide television that all homes henceforth would be subject to warrantless entry, individuals thereafter might not in fact entertain any actual expectation of privacy regarding their homes, papers, and effects.... In such circumstances, where an individual’s subjective expectations had been “conditioned” by influences alien to well-recognized Fourth Amendment freedoms, those subjective expectations obviously could play no meaningful role in ascertaining what the scope of Fourth Amendment protection was. In determining whether a “legitimate expectation of privacy” existed in such cases, a normative inquiry would be proper.
Smith v. Maryland, 442 U.S. 735, 740 n. 5, 99 S.Ct. 2577, 61 L.Ed.2d 220 (1979).
Pervasively imposing an intrusive search regime as the price of pre-trial release, just like imposing such a regime outright, can contribute to- the downward ratchet of privacy expectations. While government may sometimes condition benefits on waiver of Fourth Amendment rights — for instance, when dealing with contractors, see Zap v. United States, 328 U.S. 624, 628, 66 S.Ct. 1277, 90 L.Ed. 1477 (1946); Yin v. California, 95 F.3d 864, 872 (9th Cir.1996) (“It is clear that a contract may under appropriate circumstances diminish (if not extinguish) legitimate expectations of privacy.”), or paying welfare benefits, see Wyman v. James, 400 U.S. 309, 317-18, 91 S.Ct. 381, 27 L.Ed.2d 408 (1971) —its power to do so is not unlimited.
Government employees, for example, do not waive their Fourth Amendment rights simply by accepting a government job; searches of government employees must still be reasonable. See Nat'l Treasury Employees Union v. Von Raab, 489 U.S. 656, 665, 109 S.Ct. 1384, 103 L.Ed.2d 685 (1989) (citing O’Connor v. Ortega, 480 U.S. 709, 717, 107 S.Ct. 1492, 94 L.Ed.2d 714 (1987) (plurality opinion); id. at 731, 107 S.Ct. 1492 (Scalia, J., concurring in the judgment)). Von Raab also forecloses a milder version of the waiver theory, which would hold that an employee keeps his Fourth Amendment rights but makes all searches reasonable through his consent: The employee’s assent is merely a relevant factor in determining how strong his expectation of privacy is, see id. at 672 n. 2, 109 S.Ct. 1384, and thus may contribute to a finding of reasonableness. See page 897 infra.
The government is obviously subject to no fewer constraints when acting as sovereign than as employer, and deciding whether someone charged with a crime *893will be incarcerated before a determination of guilt is unquestionably a sovereign prerogative. “[0]ne who has been released on pretrial bail does not lose his or her Fourth Amendment right to be free of unreasonable seizures,” Cruz v. Kauai County, 279 F.3d 1064, 1068 (9th Cir.2002), and we have previously held that probationers (a group more readily subject to restrictions than pre-trial releasees, see pages 897-98 infra) do not waive their Fourth Amendment rights by agreeing, as a condition of probation, to “submit [their] person and property to search at any time upon request by a law enforcement officer.” United States v. Consuelo-Gonzalez, 521 F.2d 259, 261 (9th Cir.1975) (en banc); see also id. at 262 (“[A]ny search made pursuant to the condition included in the terms of probation must necessarily meet the Fourth Amendment’s standard of reasonableness.”).
Therefore, Scott’s consent to any search is only valid if the search in question (taking the fact of consent into account) was reasonable. To this inquiry we now turn.
2. Usually, Fourth Amendment reasonableness means that a search or seizure must be supported by probable cause, though pat-downs and similar minor intrusions need only be supported by reasonable suspicion. See Terry v. Ohio, 392 U.S. 1, 88 S.Ct. 1868, 20 L.Ed.2d 889 (1968); United States v. Brignoni-Ponce, 422 U.S. 873, 881-82, 95 S.Ct. 2574, 45 L.Ed.2d 607 (1975). But we relax these requirements “when ‘special needs, beyond the normal need for law enforcement,”’ make an insistence on the otherwise applicable level of suspicion “ ‘impracticable.’ ” Griffin v. Wisconsin, 483 U.S. 868, 873, 107 S.Ct. 3164, 97 L.Ed.2d 709 (1987) (quoting New Jersey v. T.L.O., 469 U.S. 325, 351, 105 S.Ct. 733, 83 L.Ed.2d 720 (1985) (Blackmun, J., concurring in the judgment)); City of Indianapolis v. Edmond, 531 U.S. 32, 37, 121 S.Ct. 447, 148 L.Ed.2d 333 (2000). Thus, when probable cause would normally be required, "special needs" can justify searches based on less, see id.; Vernonia Sch. Dist. 47J v. Acton, 515 U.S. 646, 653, 664-65, 115 S.Ct. 2386, 132 L.Ed.2d 564 (1995) (suspicionless drug testing of student athletes). When reasonable suspicion would normally be required, special needs may justify suspicionless seizures, see Edmond, 531 U.S. at 39, 121 S.Ct. 447 (citing Mich. Dep't of State Police v. Sitz, 496 U.S. 444, 110 S.Ct. 2481, 110 L.Ed.2d 412 (1990) (sobriety checkpoint), as such a case).
The government argues here that searching pre-trial releasees by testing them for drugs serves two special needs: (1) protecting the community from criminal defendants released pending trial and (2) ensuring that defendants show up at trial. But-at most-only the second of these claimed needs is, as the special needs exception requires, "beyond the normal need for law enforcement."
Two recent Supreme Court cases ilus-trate this important limitation on the special needs doctrine. In Edmond, the Court invalidated a roadside checkpoint program aimed at enforcing drug laws through drug-sniffing dogs and visual inspection of cars. See 531 U.S. at 35, 121 S.Ct. 447. The Court started with the observation that the suspicionless checkpoint stops were Fourth Amendment seizures requiring individualized suspicion. See id. at 40-41, 121 S.Ct. 447. It noted that it had "never approved a checkpoint program whose primary purpose was to detect evidence of ordinary criminal wrongdoing." Id. at 41, 121 S.Ct. 447. Rather, suspicionless checkpoint stops are constitutional only if their primary purpose is separate from the "general interest in crime control." Id. (quoting Delaware v. *894Prouse, 440 U.S. 648, 659 n. 18, 99 S.Ct. 1391, 59 L.Ed.2d 660 (1979)) (internal quotation marks omitted). Programs designed to secure the border, see United States v. Martinez-Fuerte, 428 U.S. 543, 556-57, 96 S.Ct. 3074, 49 L.Ed.2d 1116 (1976), or promote highway safety, see Sitz, 496 U.S. at 451, 110 S.Ct. 2481, are thus different from programs whose purpose is to deter and punish violations of ordinary criminal laws.
For much the same reason, in Ferguson v. City of Charleston, 532 U.S. 67, 121 S.Ct. 1281, 149 L.Ed.2d 205 (2001), the Court invalidated a state hospital’s practice of testing pregnant women for cocaine and providing the results to the police. The Court had upheld suspicionless drug testing programs before, but in those cases, “the ‘special need’ ... was one divorced from the State’s general interest in law enforcement.” Id. at 79, 121 S.Ct. 1281; see also id. at 77, 121 S.Ct. 1281 (citing Skinner v. Ry. Labor Executives’ Ass’n, 489 U.S. 602, 109 S.Ct. 1402, 103 L.Ed.2d 639 (1989) (drug testing of railroad employees to prevent railway accidents); Von Raab, 489 U.S. at 656, 109 S.Ct. 1384 (drug testing of Customs employees to ensure their integrity and physical fitness); Vernonia, 515 U.S. at 646, 115 S.Ct. 2386 (drug testing of student athletes to maintain order in schools)). In Ferguson, however, “the central and indispensable feature of the policy from its inception was the use of law enforcement to coerce the patients into substance abuse treatment.” Id. at 80, 121 S.Ct. 1281. The Court considered the government’s argument that the “ultimate purpose” of the testing program was the “beneficent” goal of “protecting the health of both mother and child,” but nonetheless concluded that “the purpose actually served ... ‘is ultimately indistinguishable from the general interest in crime control.’ ” Id. at 81, 121 S.Ct. 1281 (quoting Edmond, 531 U.S. at 44, 121 S.Ct. 447).
Edmond’s and Ferguson’s focus on “primary” or “ultimate” purposes requires us to examine the various possible purposes of the search here and determine which are primary. Because the subjective intent of the officers carrying out the search generally plays no role in assessing its constitutionality, see Whren v. United States, 517 U.S. 806, 813, 116 S.Ct. 1769, 135 L.Ed.2d 89 (1996), special needs analysis calls for an inquiry into “programmatic purposes,” see Edmond, 531 U.S. at 45-47, 121 S.Ct. 447; Ferguson, 532 U.S. at 81, 121 S.Ct. 1281 (“In looking to the programmatic purpose, we consider all the available evidence in order to determine the relevant primary purpose.”).
The government’s first identified purpose, protecting the community, presumably means protecting it from the criminal activities of pre-trial releasees. See Nev.Rev.Stat. § 178.4851(2); see also id. § 178.4853(9) (listing “[t]he likelihood of more criminal activity by [the releasee] after he is released” as one factor to be considered before release without bail). The dissent points out correctly that the “ ‘government’s interest in preventing crime by arrestees is both legitimate and compelling,’ ” dissent at 907 (quoting Salerno, 481 U.S. at 749, 107 S.Ct. 2095). But the government’s interest in preventing crime by anyone is legitimate and compelling. See, e.g., United States v. Restrepo, 946 F.2d 654, 674 (9th Cir.1991) (en banc) (Norris, J., dissenting) (“It goes without saying that the government has a compelling interest in protecting the community from crime.”). Crime prevention is a *895quintessential general law enforcement purpose and therefore is the exact opposite of a special need.
The second purpose, ensuring that pretrial releasees appear in court, fares somewhat better: While it has a law enforcement component — a defendant’s failure to appear in court when ordered to do so is a criminal offense, see Nev.Rev.Stat. § 199.335 — it also implicates the efficient functioning and integrity of the judicial system, cf. State v. Ullring, 741 A.2d 1065, 1068 (Me.1999), a purpose separate from the general interest in crime control.
We assume for purposes of our analysis that the non-law-enforcement purpose— the interest in judicial efficiency — is “primary” in this case. It remains to determine whether this need “is important enough to override the individual’s acknowledged privacy interest! and] sufficiently vital to suppress the Fourth Amendment’s normal requirement of individualized suspicion,” Chandler v. Miller, 520 U.S. 305, 318, 117 S.Ct. 1295, 137 L.Ed.2d 513 (1997) (striking down a program of suspicionless drug testing of candidates for public office because this showing had not been made), and then weigh it against “the intrusion on the individual’s interest in privacy,” Ferguson, 532 U.S. at 78, 121 S.Ct. 1281.
To begin with, the connection between the object of the test (drug use) and the harm to be avoided (nonappearance in court) is not obvious. The defendant may use drugs while on pre-trial release and be so overcome by the experience — he’s in a drug-induced stupor or in a hospital emergency room — that he misses his court date. Or, having made it to court, he may be too mentally impaired to participate meaningfully in the proceedings. These are conceivable justifications, but they strike us as highly unlikely. The government has produced nothing to suggest these are common enough problems to justify intruding on the privacy rights of every single defendant out on pre-trial release. Drug use during pre-trial release may also result in a defendant’s general unreliability or, more nefariously, an increased likelihood of absconding. Whether this is plausible depends on whether drug use is a good predictor of these harms- — a case that must be established empirically by the government that seeks to impose the drug testing condition.
The Supreme Court has criticized assertions of special needs based on “hypothetical” hazards that are unsupported by “any indication of a concrete danger demanding departure from the Fourth Amendment’s main rule.” Chandler, 520 U.S. at 319, 117 S.Ct. 1295. “A demonstrated problem” of drug use leading to nonappearance “would shore up” the government’s assertion of a special need. See id. As far as we can tell, the Nevada legislature has not taken *896the position that drug use among pre-trial releasees substantially impairs their tendency to show up in court; instead, it has largely left appropriate release conditions to be determined in individual cases. See Nev.Rev.Stat. §§ 178.484-.4853. Nor are courts instructed to limit their consideration to the non-law-enforcement purposes that might justify special needs searches: Release conditions may both “protect the health, safety and welfare of the community and ... ensure that [the releasee] will appear at all times and places ordered by the court.” See id. § 178.4851(2) (emphasis added); see also id. § 178.4853(7)-(9). We are thus unable to conclude that the search regime to which Scott was subjected was necessary to ensure his appearance at trial.
We are especially reluctant to indulge the claimed special need here because Scott’s privacy interest in his home, where the officers came to demand the urine sample, is at its zenith. “[P]rivate residences are places in which the individual normally expects privacy free of governmental intrusion not authorized by a warrant, and that expectation is plainly one that society is prepared to recognize as justifiable.” United States v. Karo, 468 U.S. 705, 714, 104 S.Ct. 3296, 82 L.Ed.2d 530 (1984); see also Payton v. New York, 445 U.S. 573, 589, 100 S.Ct. 1371, 63 L.Ed.2d 639 (1980) (“The Fourth Amendment protects the individual’s privacy in a variety of settings. In none is the zone of privacy more clearly defined than when bounded by the unambiguous physical dimensions of an individual’s home.... ”). Unlike public school students, who have limited privacy interests because of the state’s special custodial role, see Bd. of Educ. of Indep. Sch. Dist. No. 92 v. Earls, 536 U.S. 822, 830, 122 S.Ct. 2559, 153 L.Ed.2d 735 (2002), Customs employees, who occupy sensitive government positions, see Von Raab, 489 U.S. at 669, 109 S.Ct. 1384, or drivers and railway employees, whose activities impose safety risks on others, see Sitz, 496 U.S. at 451, 110 S.Ct. 2481; Skinner, 489 U.S. at 620, 109 S.Ct. 1402, pre-trial releasees are ordinary people who have been accused of a crime but are presumed innocent. We have already noted that Scott’s assent to his release conditions does not by itself make an otherwise unreasonable search reasonable, see section 1 supra; to the extent his assent decreased his reasonable expectation of privacy, we hold that the decrease was insufficient to eliminate his expectation of privacy in his home.
Griffin, where the Supreme Court upheld the search of a probationer’s home without probable cause, is not to the contrary. Griffin was on probation rather than pre-trial release. See 483 U.S. at 870, 107 S.Ct. 3164. The Court wrote that “[a] State’s operation of a probation system, like its operation of a school, government office or prison,” presents special needs, id. at 873-74, 107 S.Ct. 3164, and that the goals of probation would be disrupted by a warrant or probable cause requirement, id. at 875-80, 107 S.Ct. 3164. But pre-trial releasees are not probationers. “Probation, like incarceration, is ‘a form of criminal sanction imposed by a court upon an offender after verdict, finding, or plea of guilty.’ ” Id. at 874, 107 S.Ct. 3164 (quoting G. Killinger et al., Probation and Parole in the Criminal Justice System 14 (1976)). Years later, the Ferguson Court explained that “Griffin is properly read as limited by *897the fact that probationers have a lesser expectation of privacy than the public at large.” 532 U.S. at 80 n. 15, 121 S.Ct. 1281. People released pending trial, by contrast, have suffered no judicial abridgment of their constitutional rights.
Because the government has failed to make the requisite special needs showing, the police needed probable cause to test Scott for drugs. We thus cannot validate Scott’s search under the special needs doctrine.
3. Nor was the search reasonable under a more general “totality of the circumstances” approach. Scott’s position was in some ways similar to that of the probationer whose reasonable-suspicion search was upheld on this theory in United States v. Knights, 534 U.S. 112, 122 S.Ct. 587, 151 L.Ed.2d 497 (2001). But in upholding that search, the Supreme Court stressed Knights’s status as a probationer:
[T]he reasonableness of a search is determined “by assessing, on the one hand, the degree to which it intrudes upon an individual’s privacy and, on the other, the degree to which it is needed for the promotion of legitimate governmental interests.” Knights’ status as a probationer subject to a search condition informs both sides of that balance.
Id. at 118-19, 122 S.Ct. 587 (quoting Wyoming v. Houghton, 526 U.S. 295, 300, 119 S.Ct. 1297, 143 L.Ed.2d 408 (1999)). The reasoning from Griffin was thus prominently on display in Knights. On the privacy side, probationers have sharply reduced liberty and privacy interests: Probation is a form of criminal punishment, so “probationers ‘do not enjoy “the absolute liberty to which every citizen is entitled.’”” Id. at 119, 122 S.Ct. 587 (quoting Griffin, 483 U.S. at 874, 107 S.Ct. 3164 (quoting Morrissey v. Brewer, 408 U.S. 471, 480, 92 S.Ct. 2593, 33 L.Ed.2d 484 (1972))). Though the Knights Court declined to address the consent rationale, see id. at 118, 107 S.Ct. 3164; see also section 1 supra, it did reason that Knights’s signature on a form purporting to authorize searches without a warrant or “reasonable cause” as a condition of probation, reduced his reasonable expectation of privacy, id. at 114, 119-20, 122 S.Ct. 587. The government, for its part, has an enhanced interest in surveillance and control because “ ‘the very assumption of ... probation’ is that the probationer ‘is more likely than the ordinary citizen to violate the law,’ ” id. at 120, 122 S.Ct. 587 (quoting Griffin, 483 U.S. at 880, 107 S.Ct. 3164), and probation is also concerned with reintegrating the probationer into the community, see id. at 120-21, 107 S.Ct. 3164.
The dissent’s inability to see a “constitutionally relevant” distinction, see dissent at 907, between someone who has been convicted of a crime and someone who has been merely accused of a crime but is still presumed innocent, overlooks both common sense and our caselaw. Recently, in Kincade, a plurality of this court noted “the well-established principle that parolees and other conditional releasees are not entitled to the full panoply of rights and protections possessed by the general public.” 379 F.3d at 833 (plurality opinion). It stressed the “transformative changes *898wrought by a lawful conviction and accompanying term of conditional release,” id. at 834, and the “severe and fundamental disruption in the relationship between the offender and society, along with the government’s concomitantly greater interest in closely monitoring and supervising conditional releasees,” occasioned by a conviction and imposition of release conditions, id. at 835.
For our purposes, the lesson of Knights and Kincade is the same as that of Griffin: Probationers are different. Like Knights, Scott had a reduced expectation of privacy because he had signed a form that, on its face, explicitly waived the warrant requirement and implicitly (through the use of the word “random”) waived the probable cause requirement for drug testing. But Scott, far from being a post-conviction conditional releasee, was out on his own recognizance before trial. His privacy and liberty interests were far greater than a probationer’s. Moreover, the assumption that Scott was more likely to commit crimes than other members of the public is contradicted by the presumption of innocence: While it is true that the Supreme Court has upheld the constitutionality of pre-trial detention on grounds of dangerousness, the Court stressed that the statute it was upholding contained important safeguards, including the requirements that the defendant be accused of a particularly serious crime and that dangerousness be proved to a neutral judicial officer by clear and convincing evidence. See Salerno, 481 U.S. at 747, 750-52, 107 S.Ct. 2095; cf. United States v. Kills Enemy, 3 F.3d 1201, 1203 (8th Cir.1993) (contrasting pre-trial releas-ees with convicted persons awaiting sentence, and noting that the latter are “no longer entitled to a presumption of innocence or presumptively entitled to [their] freedom”). Further, the government has no concern with integrating people like Scott, who has never left the community, back into the community. The government’s interests in surveillance and control as to a pre-trial releasee are thus considerably less than in the case of a probationer. A search of Scott or his house on anything less than probable cause is not supported by the totality of the circumstances.
4. The government concedes that there was no probable cause to test Scott for drugs. Therefore, Scott’s drug test violated the Fourth Amendment. Probable cause to search Scott’s house did not exist until the drug test came back positive. The validity of the house search, which led to both the shotgun and Scott’s statement about the shotgun, is derivative of the initial drug test. That search is likewise invalid; its fruits must be suppressed.
We AFFIRM the district court’s order granting Scott’s motion to suppress.
BYBEE, Circuit Judge,
dissenting:
The majority holds that probable cause is required for the warrantless search of the person or home of a pretrial releasee. This is not only an issue of first impression in our circuit, it is an issue of first impression in any federal court. In a case to which no state is a party, the majority casually adopts a rule that carries monumental implications for the pretrial procedures employed by every state in our circuit, as well as the United States.
*899While the question is one of first impression, we are not without guidance. There is a body of jurisprudence — both state and federal — examining the status of probationers, parolees, and presentence and pretrial releasees. Based on my reading of the cases in this arena, I cannot agree with the majority’s new per se rule. If the “touchstone of our analysis under the Fourth Amendment is always ‘the reasonableness in all the circumstances of the particular governmental invasion[,]’ ” then an individualized inquiry must be undertaken to determine if the particular condition imposed is, in fact, “reasonable.” Pennsylvania v. Mimms, 434 U.S. 106, 108-09, 98 S.Ct. 330, 54 L.Ed.2d 331 (1977) (quoting Terry v. Ohio, 392 U.S. 1, 19, 88 S.Ct. 1868, 20 L.Ed.2d 889 (1968)) (emphasis added).
Accordingly, I would resolve this case by specifically examining the facts and circumstances applicable to Scott’s pre-trial release, then weighing the legitimate interests of the state against the individual privacy interests at stake. Under this approach — a familiar approach employed for warrantless searches of probationers, parolees, and presentence releasees — I do not believe that probable cause was necessary to search Scott’s person for drugs. Once obtained, the positive drug test result gave rise to probable cause to search Scott’s living room. I respectfully dissent.
I. FACTS AND PROCEEDINGS BELOW
The individualized inquiry that I find necessary calls for a fuller development of the facts than that proffered by the majority. Raymond Lee Scott was charged in Douglas County, Nevada with one felony and two misdemeanors related to the possession of methamphetamine and drug paraphernalia. Two days later, Scott was released on his own recognizance (“OR”), subject to his consent to several conditions of pre-trial release.
On a printed release form, the Nevada court checked some, but not all, of the conditions on the form. Scott agreed to submit to “random drug and alcohol testing, anytime of the day or night by any peace officer without a warrant”; and to submit himself, his residence and his vehicle “to search and seizure by any peace officer anytime day or night without a warrant for C/S [controlled substances and] ALCOHOL.” He also promised not to “carry or possess any firearms or dangerous weapons[.]” Finally, Scott’s release form provided that “[a]ny law enforcement officer having probable cause to believe the named defendant has violated a condition of this release is ordered to arrest the person.”
Shortly after his release, Douglas Swalm, a probation officer working for the Department of Alternative Sentencing, received information that Scott had in his possession a 9mm handgun, a sawed off shotgun and paraphernalia specific to the manufacture of methamphetamine. Based on this information, and without obtaining a warrant, Swalm conducted a “compliance visit” at Scott’s residence, accompanied by probation officer Nathan Almeida and several sheriffs deputies and narcotics agents. Upon their arrival, Scott invited the officers inside, where Almeida administered a urine drug test that indicated that Scott had been using methamphetamine. Scott was then handcuffed, seated on a couch in his living room, told that the officers were going to search his residence, and questioned as to whether there were any weapons in the house. Scott denied having any firearms, but admitted to having “several toy guns that his children used to play Cowboys and Indians with.”
Later, Almeida asked Scott where the “toy guns” were. Scott gestured to the *900television set across the room, where Al-meida spotted a nylon holster that appeared to have a gun in it with both the grip and the barrel protruding from the holster. Almeida testified that once Scott pointed out the gun, he could recognize it immediately. The holster contained a sawed-off .410 gauge shotgun. The officers also found a box of shotgun shells adjacent to the television.
Based on the results of the search, a grand jury charged Scott with possessing a shotgun in violation of 26 U.S.C. §§ 5841, 5861(d) and 5871. Scott moved to suppress the shotgun and the statements he made during the search. The government conceded that the search was not supported by probable cause, but argued that Swalm reasonably suspected that Scott was in violation of the conditions of his release.
The district court granted Scott’s motion to suppress. Balancing Scott’s privacy interests against the “legitimate interests of the state in light of the totality of the circumstances,” the district court concluded that probable cause was needed to search Scott’s home. United States v. Scott, CR-N-03-0122-DWH, Order (D.Nev., Jan. 26, 2004) (citing United States v. Knights, 534 U.S. 112, 118-19, 122 S.Ct. 587, 151 L.Ed.2d 497 (2001)). Finding no probable cause, the court held the search unreasonable in violation of the Fourth Amendment. The government timely appealed to this court.
II. STANDARD FOR WARRANTLESS SEARCH OF PRETRIAL RELEASEE
Scott contends that the district court properly determined that probable cause, rather than reasonable suspicion, was necessary to justify the search of his home. The government maintains that reasonable suspicion is the only standard applicable to pretrial releasees, and that such suspicion was satisfied in this case.
Although the question is one of first impression, we do not write on a clean slate. Accordingly, a detailed examination of the body of Fourth Amendment case law exploring the status of individuals with diminished liberty interests is useful to understanding the precise nature of the liberty interests under scrutiny here.
A. Fourth Amendment Status of Individuals tuith Diminished Liberty Interests
1. Probationers, Parolees, and Presen-tence and Pretrial Releasees
Courts have distinguished the liberty interests of individuals on probation and parole from ordinary citizens who have not been convicted of any crime. Warrantless searches that would not meet constitutional standards if other persons were the targets often meet constitutional muster when the target of the search is a parolee or probationer. United States v. Consuelo-Gonzalez, 521 F.2d 259, 259 (9th Cir.1975). We have held that warrantless searches of parolees and probationers are acceptable provided that they are conducted pursuant to the terms of the probation, United States v. Richardson, 849 F.2d 439 (9th Cir.1988), or they are conducted to advance the goals of the individual’s probation, rather than merely for the purposes of investigation, United States v. Ooley, 116 F.3d 370, 372 (9th Cir.1997). Probationers and parolees’ homes are often searched without a warrant or probable cause. See, e.g., United States v. Tucker, 305 F.3d 1193 (10th Cir.2002) (allowing search of parolee’s home based on reasonable suspicion); Owens v. Kelley, 681 F.2d 1362 (11th Cir.1982) (upholding probation condition stating that defendant convicted of drug possession must submit to search *901of his home at any time whenever requested by a probation officer); Latta v. Fitzharris, 521 F.2d 246, 250 (9th Cir.1975) (“the parolee and his home are subject to search by the parole officer when the officer reasonably believes that such search is necessary in the performance of his duties”); People v. Reyes, 19 Cal.4th 743, 80 Cal.Rptr.2d 734, 968 P.2d 445 (1998) (upholding search condition as it applied to parolee); Wayne R. Lafave, Search And SeizuRE: A Treatise On The FoüRth Amendment § 10.10 (4th ed.2004).
Similarly, courts have determined that individuals who have been convicted of crimes but remain free on bond or OR pending sentencing have a diminished expectation of privacy as compared to ordinary citizens. Courts have determined that probable cause is not necessary to search the homes of convicts released pre-sentencing. In State v. Fisher, 145 Wash.2d 209, 35 P.3d 366 (2001), the Washington Supreme Court held that although probable cause is normally required for issuance of a warrant, convicts released pending sentencing may be arrested on the basis of a “reasonable grounds” standard. The court distinguished the “due process rights of an accused [from those of] a defendant who has already been adjudged ‘guilty[.]’ ” Id. at 375. The court further pointed out that:
Both circumstances raise different expectations of privacy and levels of constitutional protections. An accused’s liberty is subject to restraint through an arrest and the jurisdiction of the courts. A convicted and sentenced felon is subject to the jurisdiction of the Department of Corrections. A convicted felon who awaits sentencing is still subject to the court’s jurisdiction, but yet does not possess the same constitutional rights as one merely accused.... Accordingly, [the defendant’s] rights must be analyzed not from the status of an accused person, but from her status as a convicted felon released on personal recognizance and awaiting sentencing.
Id. at 375-76 (emphasis in original). Other courts have similarly observed that individuals released after conviction, but not yet on probation or parole, have a “reduced expectation of privacy[,]” and thus, “the police needed only a reasonable basis to conduct a warrantless search of their home.” See State v. Anderson, 140 Idaho 484, 95 P.3d 635, 638 (2004).
We have drawn this same distinction. In Portillo v. U.S. Dist. Court for the Dist. of Ariz., 15 F.3d 819, 823-24 (9th Cir.1994), we held that the supervisory nature of postconviction, presentence release is a “special need” justifying drug tests on the basis of reasonable suspicion rather than probable cause. In balancing the defendant’s privacy interests against the state’s supervisory interests, the court specifically considered the fact that the defendant “[had] been convicted of theft and [was] awaiting sentencing.” Id.
A few courts have considered the Fourth Amendment status of individuals, like Scott, who have been released pretrial on bail or on their own recognizance. We have commented only briefly on the status of pretrial releasees. In United States v. Kills Enemy, 3 F.3d 1201 (8th Cir.1993), we contrasted an individual on pretrial release with a convicted person awaiting sentence in that the latter “is no longer entitled to a presumption of innocence or presumptively entitled to his freedom.” Id. at 1203. And, in Cruz v. Kauai County, 279 F.3d 1064 (9th Cir.2002), we observed that “one who has been released on pretrial bail does not lose his or her Fourth Amendment right to be free of unreasonable seizures.” Id. at 1068.
In contrast, individuals confined in prison pending trial have no greater privacy *902rights than other prisoners. Bell v. Wolfish, 441 U.S. 520, 99 S.Ct. 1861, 60 L.Ed.2d 447 (1979) (treating pretrial detainees the same as prisoners convicted of an offense); William E. Ringel, Seaeches AND SEIZURES, ARRESTS AND CONFESSIONS § 17:9 (2005). Thus, a search of a pretrial detainee’s cell, Soldal v. Cook County, 506 U.S. 56, 65, 113 S.Ct. 538, 121 L.Ed.2d 450 (1992); Hudson v. Palmer, 468 U.S. 517, 104 S.Ct. 3194, 82 L.Ed.2d 393 (1984), as well as the random monitoring and recording of a pre trial detainee’s telephone conversations on a prison telephone have not been held to violate the Fourth Amendment. See United States v. Willoughby, 860 F.2d 15 (2d Cir.1988).
Although pretrial releasees may not lose their Fourth Amendment rights, until today’s decision we have not squarely addressed whether pretrial releasees have diminished Fourth Amendment rights. Specifically, does a search of a pretrial releasee require probable cause? But the question has been addressed by state courts. Each has determined that war-rantless search conditions may be imposed on a pretrial releasee so long as those conditions are reasonable. See, e.g., State v. Ullring, 741 A.2d 1065, 1069 n. 3, 1073 (Me.1999); In re York, 9 Cal.4th 1133, 40 Cal.Rptr.2d 308, 892 P.2d 804 (1995).
The leading case is In re York, where the California Supreme Court determined that a state statute requiring compliance with “reasonable conditions” on pretrial release allowed for warrantless searches and random drug testing. Unable to post the bail prescribed for their offenses, the petitioners in York were given the choice of remaining in custody pending trial upon the charges or obtaining OR release. To obtain OR release, they had to agree to specified conditions, including a requirement that they “[s]ubmit to drug [and, in some instances, alcohol] testing” and “[p]ermit search and seizure of his/her person, residence, and vehicle by any peace officer without a search warrant.” Id. at 806.
Petitioners argued that the imposition of warrantless drug testing and search conditions upon OR releasees violated the Fourth Amendment. Id. The court rejected this claim for two reasons. The court first pointed out that defendants released on OR lack
the same reasonable expectation of privacy as that enjoyed by persons not charged with any crime, and by defendants who have posted reasonable bail. Unlike persons in these latter categories, however, a defendant who is unable to post reasonable bail has no constitutional right to be free from confinement prior to trial and therefore lacks the reasonable expectation of privacy possessed by a person unfettered by such confinement. Because an incarcerated individual generally is subject to random drug testing and warrantless search and seizure in the interest of prison security, the conditions challenged in the present case do not place greater restrictions upon an OR releasee’s privacy rights than the releasee would have experienced had he or she not secured OR release. Viewed from this perspective, the challenged conditions do not require an OR releasee to “waive” Fourth Amendment rights that he or she would have retained had OR release been denied. Instead, the conditions simply define the degree of liberty that the court or magistrate, in his or her discretion, has determined to grant to the OR re-leasee.
Id. at 813. Second, the court stated that the conditions are not unconstitutional because “a pretrial detainee is not required to agree to such restrictions, but rather is subject to them only if he or she consents *903to their imposition, in exchange for obtaining OR release.” Id. at 814. The court reasoned that “one who otherwise would be incarcerated prior to judgment[because he cannot post bail] is offered the opportunity to obtain OR release^] he or she is not entitled to unconditional bail-free release, but may obtain OR release only in the discretion of the court or magistrate, and only upon those reasonable conditions attached to the release.” Id. at 814. Recognizing that such restrictions are not unlimited, the court pointed out that the conditions must be “reasonable under the circumstances,” which in turn depends on “the relationship of the condition to the crime or crimes with which the defendant is charged and to the defendant’s background, including his or her prior criminal conduct.” Id. at 815 n. 10. The court then held the search and drug testing conditions reasonable under the circumstances because they “clearly relate to the prevention and detection of further crime and thus to the safety of the public.” Id. at 810. Accordingly, the court concluded that warrantless searches and random drug testing did not violate petitioners’ Fourth Amendment rights.
In State v. Ullring, 741 A.2d 1065 (Me. 1999), the Supreme Court of Maine reached the same decision, concluding that probable cause is not required to search the home of a pretrial releasee. In Ullring, a defendant was arrested for possessing marijuana and other drug paraphernalia. Id. at 1066. He posted bail and signed a bail bond requiring “him to submit to random searches of his person, residence, and vehicle.” Id. After a war-rantless search of his home uncovered marijuana, the defendant challenged the trial court’s denial of his motion to suppress the evidence.
The court began by considering the objectives of the bail system in Maine, observing that its purpose is “to ensure the appearance of the defendant at trial and to do so without incarceration as long as conditions can be imposed which will fulfill that purpose and the purpose of ensuring the integrity of the judicial process.” Id. at 1072. The court reasoned that
[b]ail conditions, such as a prohibition against possession of illegal drugs and searches for illegal drugs, help to ensure that defendants whose backgrounds and charges indicate that substance abuse is a significant problem will show up at court. It is reasonable to expect that a defendant who maintains sobriety is more likely to appear in court on the appointed dates than a defendant who is under the influence of drugs or alcohol.
Id. at 1072-73. The court concluded that the random search condition was constitutional because the condition was reasonable as applied to the “history and personal situation of the defendant.” Id. at 1073. For support, the court appealed to the Supreme Court’s seminal decisions in Griffin and Knights.
*904In Griffin, the Court upheld a Wisconsin law that permitted a probation officer to search a probationer’s home without a warrant as long as his supervisor approved and as long as there were “reasonable grounds” to believe that contraband was present. Griffin, 483 U.S. at 870-71, 107 S.Ct. 3164. Based on a tip, a probation officer searched Griffin’s home and found a firearm. Griffin was then charged with being a felon in possession of a firearm. Id. at 872, 107 S.Ct. 3164. The Wisconsin Supreme Court had previously announced a per se rule that “a probation officer may, consistent with the Fourth Amendment, search a probationer’s home without a warrant, and with only ‘reasonable grounds’ (not probable cause) to believe that contraband is present.” Id. at 872, 107 S.Ct. 3164. The Court held that it was “unnecessary to embrace a new principle of law ... that any search of a probationer’s home by a probation officer satisfies the Fourth Amendment as long as the information possessed by the officer satisfies a federal ‘reasonable grounds’ standard,” but it was enough to say that “this warrantless search did not violate the Fourth Amendment.” Id. (emphasis added). The court concluded that “the search of Griffin’s home satisfied the demands of the Fourth Amendment because it fulfilled a ‘special need[, which] make the warrant and probable-cause requirement impracticable.’ ” Id. (citing New Jersey v. T.L.O., 469 U.S. 325, 105 S.Ct. 733, 83 L.Ed.2d 720 (1985)).
The Court reexamined the issue in Knights, and held that a warrantless search of a probationer’s apartment was reasonable under the Fourth Amendment where it was authorized by a condition of his probation and supported by reasonable suspicion. Knights, 534 U.S. at 122, 122 S.Ct. 587. After obtaining probation for a misdemeanor drug offense, Knights signed a condition that he would “[sjubmit his ... person, property place of residence, vehicle, personal effects, to search at anytime, with or without a search warrant, warrant of arrest or reasonable cause by any probation officer or law enforcement officer.” Id. at 114, 122 S.Ct. 587. While Knights was on probation, a police officer conducted an investigatory search of his home and found materials indicating that he participated in a conspiracy to commit arson. Id. at 116, 122 S.Ct. 587.
Knights argued that “a warrantless search of a probationer satisfies the Fourth Amendment only if it is ... ‘a special needs’ search conducted by a probation officer.” Id. at 117, 122 S.Ct. 587. The Court rejected this rationale, again emphasizing that Griffin did not decide that all warrantless searches of probationers were reasonable within the meaning of the Fourth Amendment. Id. at 117-18, 122 S.Ct. 587. Instead, the Court relied on a “general Fourth Amendment approach of ‘examining the totality of the circumstances,’ Ohio v. Robinette, 519 U.S. 33, 39, 117 S.Ct. 417, 136 L.Ed.2d 347 (1996), with the probation search condition being a salient circumstance.” Knights, 534 U.S. at 118, 122 S.Ct. 587. See also United States v. Stokes, 292 F.3d 964, 967-68 (9th Cir.2002). This approach consisted of “assessing, on the one hand, the degree to which it intrudes upon an individual’s privacy and, on the other, the degree to which it is needed for the promotion of legitimate governmental interests.” Id. at 119 (citing Wyoming v. Houghton, 526 U.S. 295, 300, 119 S.Ct. 1297, 143 L.Ed.2d 408 (1999)). The probation order “clearly expressed the search condition and Knights was unambiguously informed of it.” Id. at 119, 122 S.Ct. 587. “The probation condition thus significantly diminished Knights’ reasonable expectation of privacy.” Id. The Court also recognized the state’s concern in preventing a probationer *905from engaging in further criminal activities. Id. at 120-21, 122 S.Ct. 587. Because the state may “justifiably focus on probationers in a way that it does not on the ordinary citizen,” reasonable suspicion rather than probable cause may justify a warrantless search. Id. at 121, 122 S.Ct. 587.
2. Totality of the Circumstances
Although the majority cites Griffin and Knights, it does not consider the various factors that are relevant in determining whether the search of Scott’s home and person were reasonable. In my view, a balancing approach that incorporates those factors is appropriate here. Toward this end, I consider the state’s interests, which include the purposes of the OR release and bail system in Nevada, state law standards relating to eligibility for pretrial release, and specific concerns applicable to Scott’s pre-trial release, taking into account the crime with which he was chárged and the conditions to which he consented.
The imposition of pretrial search conditions must be adapted to the pretrial re-leasee; the conditions should be related to a public purpose such as rehabilitating the offender or protecting the public. See, e.g., Owens, 681 F.2d at 1366-67; Consuelo-Gonzalez, 521 F.2d at 263. Even as I question the “categorical no” approach adopted by the majority, I would not endeavor to create a one-size-fits-all standard for approving all searches of pretrial re-leasees; this approach was rejected by the Court in Griffin with respect to probationers and it is equally ill-suited to pretrial releasees.
Finally, although the balancing test offered by the Court in Knights seems to invite a series of finely tailored standards, lower federal courts are not free to insert new standards into the gap between reasonable suspicion and probable cause. United States v. Montoya de Hernandez, 473 U.S. 531, 540-41, 105 S.Ct. 3304, 87 L.Ed.2d 381 (1985) (rejecting an intermediate “clear indication” standard as incompatible with the Fourth Amendment’s “reasonableness” approach). Thus, recognizing “reasonableness” as the hallmark of the Fourth Amendment, I apply the standards announced in Knights, balancing Scott’s privacy interests against the legitimate interests of the state in light of the totality of the circumstances. Knights, 534 U.S. at 118-19, 122 S.Ct. 587.
B. Fourth Amendment Balancing
In accord with the Court’s analysis in Griffin, I turn to an examination of the standards applied by Nevada courts in releasing Scott on his own recognizance. While “the validity of a search conducted by state law enforcement officers is ultimately a question of federal law”, federal courts “must look to state law to determine the validity of the underlying [release] condition itself, and may consider state precedent for its persuasive value[.]” Ooley, 116 F.3d at 372 (internal citations and quotations omitted). Before releasing a person without bail, a Nevada “court may impose such reasonable conditions on *906the person as it deems necessary to protect the health, safety and welfare of the community to ensure that the person will appear at all times and places ordered by the court ...” Nev. Rev. Stat. § 178.484(8). Although the Nevada statutes do not specifically mention warrant-less searches or random drug testing, the list is expressly nonexclusive. Id.
1. The State’s Interests
a. Protecting the public. The majority rejects the state’s interest in protecting the public as a “quintessential general law enforcement purpose [which is] the exact opposite of a special need,” and argues that the presumption of innocence insulates the pretrial releasee from the claim that he is “more likely to commit crimes than other members of society.” Maj. op. at 894-95, 898. First, the state’s interest is not so easily dismissed by referring to the state’s general duty to protect the public or the presumption of innocence. The Court has rejected a similar argument in Knights that warrantless searches of probationers must serve a “special need,” 534 U.S. at 117, 122 S.Ct. 587, and not a more general purpose. Moreover, while protecting its citizens is the first duty of government, see Marbury v. Madison, 5 U.S. (1 Crunch) 137, 163, 2 L.Ed. 60 (1803), Nevada’s concern for the safety of the public is not “general law enforcement” when it is manifested in pretrial conditions tailored to this defendant.
Second, the accused enjoys the presumption of innocence as a trial right; an accused does not enjoy the same presumption with respect to ordinary civil rights of citizens, such as freedom of movement. See, e.g., U.S. Const, amend. VIII (“Excessive bail shall not be required”; emphasis added).
Both courts and Congress have implicitly rejected the majority’s argument by treating persons indicted for crimes differently than ordinary citizens. In Wolfish, the Supreme Court rejected a similar argument, concluding that “[t]he presumption of innocence is a doctrine that allocates the burden of proof in criminal trials ... [b]ut it has no application to a determination of the rights of a pretrial detainee during confinement before his trial has even begun.” 441 U.S. at 533, 99 S.Ct. 1861. Likewise, in Speight v. United States, 569 A.2d 124 (D.C.Ct.App.1989), the court held that a local statute that punished crimes committed while on pretrial release greater than those committed by ordinary citizens did not violate the presumption of innocence. See also 18 U.S.C. § 922(n) (specifically punishing indicted persons who receive or ship firearms in interstate commerce); United States v. Craven, 478 F.2d 1329, 1340 (6th Cir.1973) (rejecting claims that the statutory classification punishing indicted persons *907is irrational on the ground that, in treating indicted persons differently, it “adversely affect[s] the presumption of innoeence[ ]”; stating that the classification is valid since Congress’ conclusion that fact of felony indictment is “so often indicative of a propensity for violence” is “eminently reasonable”); United States v. Brown, 484 F.2d 418, 424 (5th Cir.1973) (same); United States v. Quiroz, 449 F.2d 583, 585 (9th Cir.1971) (same). Cf. 18 U.S.C. § 3577 (approving of consideration of prior arrests in imposing penalties by providing that “[n]o limitation shall be placed on the information concerning the background, character, and conduct of a person convicted of an offense which a court of the United States may receive and consider for the purpose of imposing an appropriate penalty”). Individuals on pretrial release may be treated differently than ordinary citizens without violating the presumption of innocence.
The majority’s lack of consideration for the state’s expressed interests is especially irresponsible in cases involving “drugs or illegal weapons” where authorities supervising the convict “must be able to act based on a lesser degree of certainty than the Fourth Amendment would otherwise require in order to intervene before [the person] does damage to himself or society.” Griffin, 483 U.S. at 879, 107 S.Ct. 3164; see Kills Enemy, 3 F.3d at 1203; State v. Anderson, 140 Idaho 484, 95 P.3d 635, 638 (2004) (“Because [Defendant’s] convictions were for drug crimes, a heightened need of supervision was necessary to protect them and society”; thus, “the [Defendant’s] convictions and past drug history, combined with the rumors and reports ... of ... police and [their] neighbor are sufficient to establish reasonable grounds for the search.”).
Scott’s status as a pre-trial releasee distinguishes him from the probationer considered in Griffin, but the distinction is not constitutionally relevant. The Court’s analysis in Griffin and Knights should apply equally to the facts of this case. Scott was arrested for felony drug possession and the probation officers searched his home based on reasonable suspicion that he possessed firearms and drug paraphernalia. Perhaps the state has a lesser interest where conviction has not yet been established, but surely the state retained some interest in intervening before Scott did “damage to himself or society.” Griffin, 483 U.S. at 879, 107 S.Ct. 3164.
Moreover, “[t]he government’s interest in preventing crime by arrestees is both legitimate and compelling.” United States v. Salerno, 481 U.S. 739, 749, 107 S.Ct. 2095, 95 L.Ed.2d 697 (1987). See also Speight, 569 A.2d at 126-27 (discussing Congress’s concern with the increase in crimes committed during pretrial release). By failing to recognize these interests, the majority grossly misrepresents the government’s interest in protecting the public through supervising individuals on pretrial release.
b. Securing attendance at trial. With regard to Nevada’s second articulated interest, ensuring that the defendant appears in court, the majority hypothesizes that the state is concerned that the defendant “may use drugs ... and be so overcome by the experience ... that he misses his court date” or “may be too mentally impaired to participate meaningfully in the proceedings.” Maj. op. at 895. The majority writes this off as “highly unlikely” and asserts that the “government has produced nothing to suggest these are common enough problems to justify intruding on the privacy rights of every single defendant out on pre-trial release” and must do so “empirically.” Id.
Thus, without explanation, the majority requires that state governments empirical*908ly prove that drug use is preventing individuals from appearing in court before they can require consent to drug testing in exchange for pretrial release. See id. There are other reasons a state might link drug testing with attendance at trial. Even if the state was not concerned with physical attendance, the state has a strong interest in preserving its judicial resources. Drug testing helps ensure that the accused is physically and mentally prepared for trial, so that there are no delays or claims that the defendant was unable to understand the proceedings or participate in his defense.
Even assuming that drug use does not generally affect a pretrial releasee’s likelihood of appearing in court, the majority ignores the other state interests underlying the conditions. Although random drug tests “cannot be said to relate directly to the likelihood that a defendant will comply with his or her duty to attend subsequent court hearings,” the conditions “clearly relate to the prevention and detection of further crime and thus to the safety of the public.” In re York, 40 Cal.Rptr.2d 308, 892 P.2d at 810, 812. Requiring states to make an empirical showing before imposing a drug testing condition ties the hands of states in preventing crimes and protecting the public. Moreover, today’s holding carries monumental implications for the numerous state governments that regularly require, where drug offenses are concerned, submission to drug testing in exchange for pretrial release.
2. Scott’s Interests
As to Scott’s interest, the searches were conducted at Scott’s home, a location specially protected by the Fourth Amendment. See generally Kyllo v. United States, 533 U.S. 27, 121 S.Ct. 2038, 150 L.Ed.2d 94 (2001). No criminal judgment or sanction had been imposed on Scott at the time of the search. Further, at the time of arrest he was not carrying any dangerous weapons. Although all of these factors favor privacy protection, Scott’s initial arrest was for felony possession of methamphetamine and two misdemeanors' — possession of drugs and drug paraphernalia. Such drugs are frequently used or stored in the home. Thus, Scott’s reasonable expectation of privacy may be somewhat greater than that of a probationer, parolee, or presentence releasee, but it is less than that of an “ordinary citizen.” Moreover, Scott’s reasonable expectation of privacy is diminished somewhat by his agreement to place himself under the supervision of the Department of Alternative Sentencing and waive the warrant requirement in exchange for being released on OR. The search condition itself, a “salient circumstance” in the Fourth Amendment balance, Knights, 534 U.S. at 118, 122 S.Ct. 587, allowed warrantless searches for controlled substances and alcohol, and implicitly waived the probable cause requirement by imposing “random” drug testing. The consent form “clearly expressed the search condition” and Scott was “unambiguously informed of it.” Id. at 119, 122 S.Ct. 587. The conditions on the form, to submit to random drug testing and war-rantless searches of his home for alcohol and controlled substances, were also related to the felony and misdemeanor drug crimes with which he was charged. “The [release] condition thus significantly diminished [Scott’s] reasonable expectation of privacy.” Id.
In my view, the Knights balance tips more favorably in the direction of the state’s legitimate interests as it concerns the random drug testing condition. Scott knowingly consented to random drug testing in exchange for release from prison pending trial. I would hold that the state did not violate the Constitution by requiring Scott to submit to a random drug test *909based on reasonable suspicion. Once the state administered the drug test, and it came back positive, the officers had probable cause to arrest Scott, search his living room, and question him as to the presence of any weapons on the premises. Thus, I would hold that the guns were obtained during a lawful search. I find the Court’s decisions in Griffin and Knights instructive in this regard.
The majority attempts to distinguish Ghiffin by stating that “pre-trial releasees are not probationers,” and “[pjeople released pending trial, by contrast, have suffered no judicial abridgment of their constitutional rights.” Maj. op. at 896-97. While technically I agree that pretrial re-leasees have not had a judicial abridgment of their constitutional rights, they have a lesser expectation of privacy than an ordinary citizen. A pretrial releasee suffers great burdens and is “scarcely at liberty[.j” Albright v. Oliver, 510 U.S. 266, 279, 114 S.Ct. 807, 127 L.Ed.2d 114 (1994) (Ginsburg, J., concurring). A person facing pending charges and released on their own recognizance is “required to appear in court at the state’s command[,j ... is often subject, as in this case, to the condition that he seek formal permission from the court (at significant expense) before exercising what would otherwise be his unquestioned right to travel outside the jurisdiction.” Id. at 278, 114 S.Ct. 807. A defendant who could not post bail or obtain release on OR faces a much larger deprivation of liberty by being confined pending trial. “Pretrial confinement may imperil the suspect’s job, interrupt his source of income, and impair his family relationships.” Gerstein v. Pugh, 420 U.S. 103, 114, 95 S.Ct. 854, 43 L.Ed.2d 54 (1975). Detention may limit a defendant’s preparation for trial by limiting his access to his attorney and potential witnesses. Stack v. Boyle, 342 U.S. 1, 4, 72 S.Ct. 1, 96 L.Ed. 3 (1951). It may also “result in permanent stigma and loss of reputation to the defendant.” United States v. Motamedi, 767 F.2d 1403, 1414 (9th Cir.1985) (Boochever, J., concurring and dissenting in part). As Judge Boochever explained
The magnitude of these concerns is increased by the fact that the injuries consequent upon pretrial confinement may not be reparable upon a subsequent acquittal. Society has no mechanism to recompense an individual for income lost or damages to a career due to pretrial confinement. Nor do we compensate the individual and his family for their mental suffering and loss of reputation due to pretrial incarceration.
Id. Further, “his employment prospects may be diminished severely, he may suffer reputational harm, and he will experience the financial and emotional strain of preparing a defense.” Id. Moreover, a defendant released on his own recognizance, even though he has not been charged with a crime, is considered to be “in custody” for some purposes. Hensley v. Municipal Ct., 411 U.S. 345, 93 S.Ct. 1571, 36 L.Ed.2d 294 (1973) (pretrial releasee is considered in custody for habeas purposes); In re Floyd, 413 F.Supp. 574, 576 (D.Nev.1976) (recognizing that a defendant released on OR, but who has not been convicted and sentenced, is also considered to be “in custody”).
Importantly, the common law seems to have regarded the difference between pretrial incarceration, bail, and other ways to secure a defendant’s court attendance as different “methods of retaining control over a defendant’s person[,j which was in custody.” Albright, 510 U.S. at 277-78, 114 S.Ct. 807 (Ginsburg, J., concurring) (citing 2 M. Hale, Pleas Of The CROWN 124 (“he that is bailed, is in supposition of law still in custody, and the parties that take him to bail are in law his keepers”); 4 W. Blackstone, Commentaries 297 (bail in *910both civil and criminal cases is “a delivery or bailment, of a person to his sureties, ... he being supposed to continue in their friendly custody, instead of going to gaol [jail].”))- Thus, those complying with release conditions are able to forgo a deprivation of liberty much greater than any release condition.
This is not to say that all release conditions should be deemed constitutional. In fact, there have been several instances where courts have found release conditions too constrictive on liberty. See Evans v. Ball, 168 F.3d 856, 860-61 (5th Cir.1999) (holding that a combination of pretrial release restrictions, including restriction on the amount of interstate travel, violate the Fourth Amendment), overruled on other grounds by Castellano v. Fragozo, 352 F.3d 939 (5th Cir.2003); Gallo v. City of Phila., 161 F.3d 217, 224-25 (3d Cir.1998) (same); Murphy v. Lynn, 118 F.3d 938, 942, 946 (2d Cir.1997) (same); but see Karam v. City of Burbank, 352 F.3d 1188 (9th Cir.2003) (holding that the conditions of defendant’s “OR release — requiring that she obtain permission of the court before leaving the state and that she make court appearances — [did not] amount[ ] to a seizure under the Fourth Amendment[ ]”). However, individuals charged with a crime and released before trial are not like ordinary citizens. While “[pjretrial releasees are not probationers,” they are separated from confinement only by a few hundred dollars or a signature on a consent form. Maj. op. at 896.
This last point requires closer examination of the implications of the majority’s ruling. The majority treats Scott’s assent to the conditions of his OR release as a question of whether the Fourth Amendment permits Scott to waive his Fourth Amendment rights. This seems quite backwards to me. It seems to me that at the time Scott agreed to these conditions in exchange for release on OR he was in a much better position than we are to weigh the reasonableness of the government’s proposed course of conduct. Unless we can find some irreducible right or moral imperative within the Fourth Amendment, one that absolutely forbids pretrial detainees from agreeing to any conditions before they are released, the majority’s approach begs the question.
There are, of course, constitutionally irreducible rights — the right not to be a slave being the prime example. The Court has also suggested that government may not condition to receipt of government largesse, license, or privilege on the waiver of certain rights guaranteed by the Constitution, at least where the condition bears no plausible relationship to the receipt of the benefit. The receipt of a tax exemption cannot be conditioned, for example, on an express waiver of the privilege of criticizing the government. See, e.g., Speiser v. Randall, 357 U.S. 513, 518-19, 78 S.Ct. *9111332, 2 L.Ed.2d 1460 (1958) (“To deny[a tax] exemption to claimants who engage in certain forms of speech is in effect to penalize them for such speech[and] necessarily will have the effect of coercing the claimants to refrain from the proscribed speech.”). But no one has ever suggested that the rights of security and privacy in our “persons, houses, papers, and effects” cannot be infringed by statute or waived by agreement, at least when the infringement is related in some rational way to changes in the individual’s legal status. Griffin and Knights are conclusive evidence to the contrary. See also Wyman v. James, 400 U.S. 309, 326, 91 S.Ct. 381, 27 L.Ed.2d 408 (1971) (holding that a state that conditioned payments under its Aid to Families with Dependent Children (AFDC) program on the “recipient’s submission to warrantless searches of her home” did not unconstitutionally burden Fourth Amendment freedoms). I am not suggesting that there are no limits to what the government may demand from an OR releasee; I would hold in this case that the conditions Nevada exacted are not unreasonable.
The majority opinion may free Scott from the consequences of the state’s discovery of a sawed-off shotgun in his home, but in the end today’s opinion is not a liberty-enhancing decision. As the majority acknowledges, “many pre-trial detainees willingly consent to such conditions, preferring to give up some rights in order to sleep in their own beds while awaiting trial.” Maj. op. at 890. Today’s decision strikes down Nevada’s practice of offering pretrial detainees the option of being released on OR and sleeping in their own beds in exchange for agreeing to a limited number of conditions that the state believes will protect the public and secure the attendance of the accused at trial. But the implications of the majority’s new per se rule could hardly be more severe or far-reaching.
Every state in this circuit has a rule, similar to Nevada’s, granting state judges broad discretion in the fashioning of pretrial release conditions. See, e.g., Cal. Penal Code § 1318(a)(2) (requiring a “defendant’s promise to obey all reasonable conditions imposed by the court or magistrate” before he can be released on his own recognizance); Oe. Rev. Stat. § 135.260(l)(d) (“Conditional release may include one or more of the following conditions ... [a]ny other reasonable restriction designed to assure the defendant’s appearance.”); Nev. Rev. Stat. § 178.484(8) (permitting trial courts to “impose such reasonable conditions on the person as it deems necessary to protect the health, safety and welfare of the community to ensure that the person will appear at all times and places ordered by the court ... ”); ARiz. Rules Crim. Peoc., Rule 7.3(b)(4) (providing for pre-trial release subject to “[a]ny other condition ... which the court deems reasonably necessary”); Wash. Cr. R. 3.2(b)(7) (empowering state courts to “[ijmpose any condition other than detention deemed reasonably necessary to assure appearance as required”); Haw. Rev. Stat. § 804-7.1(9) (permitting court to require, as a condition of OR release, that the defendant “satisfy any other condition reasonably necessary to assure the appearance of the person as required and to assure the safety of any other person or community”); Alaska Stat. § 12.30.020(b)(7)-(e) (permitting courts to “impose any other condition considered reasonably necessary to assure the defendant’s appearance as required and the safety of the alleged victim, other persons, or the community”, and setting forth eleven factors the judge should take into account when fashioning conditions); Montana Stat. 46-9-108(1) (“The court *912may impose any condition that will reasonably ensure the appearance of the defendant as required or that will ensure the safety of any person or the community....”); Idaho Ceim. Rules, Rule 46(c) (allowing court to “impose such reasonable terms, conditions and prohibitions as the court finds necessary in the exercise of its discretion”). Moreover, at least one other state in this circuit has interpreted the rule to permit random drug testing and warrantless search conditions. See In re York, 40 Cal.Rptr.2d 308, 892 P.2d at 815 (upholding a California pretrial condition requiring the defendant to submit to random drug testing and war-rantless searches and seizures). See also Ullring, 741 A.2d at 1073 (concluding that a random search condition employed by trial courts in Maine did not offend the Constitution).
The United States likewise employs a rule similar to Nevada’s, granting broad discretion where pre-trial releasees are concerned. See 18 U.S.C. § 3142(c)(l)(B)(xiv) (providing for pretrial release subject to “any other condition that is reasonably necessary to assure the appearance of the person as required and to assure the safety of any other person and the community”). Importantly, the standard pre-trial release form used by federal courts across the nation requires, as a condition of release, the defendant to “submit to any method of testing required by the pretrial services office or the supervising officer for determining whether the defendant is using a prohibited substance.” 7 Fed. Proc. Forms § 20:110, Order Setting Conditions of Release, Additional Conditions of Release (AO 199B). The form continues: “Such methods may be used with random frequency and include urine testing, the wearing of a sweat patch, a remote alcohol testing system, and/or any form of prohibited substance screening or testing.” Id. (emphasis added). The majority’s decision invalidates the United States government’s pre-trial release condition unless federal officers can first demonstrate probable cause to support the drug test.
We cannot predict with certainty how the states or the United States will respond to the majority’s new per se rule prohibiting warrantless search conditions and random drug testing without probable cause. It is not hard to imagine that some jurisdictions will decide that releasing persons accused of crimes on OR without such conditions will not serve the public interest. They may respond by either insisting on bail or simply holding the accused pending trial. Those pre-trial detainees might well have “preferred] to give up some rights in order to sleep in their own beds while awaiting trial,” maj. op. at 890, but, under the majority’s decision, their Fourth Amendment rights will be secure while they rest in the county jail.
I am confident that, working within the framework of Griffin and Knights, there is little danger that we will “[g]iv[e] the gov*913ernment free rein to grant conditional benefits ... [and] abuse its power by attaching strings strategically striking lopsided deals.” Maj. op. at 891. Nothing that I have said here would sanction such actions, and I do not see that Nevada has done so here.
III. CONCLUSION
I would hold that cases involving pretrial releasees are subject to a balancing test that weighs the legitimate interests of the state against the individual privacy interests at stake in light of the unique circumstances and facts alleged. Using this approach, I conclude that the search and seizure conducted here were valid; the officers needed only reasonable suspicion to administer the drug test and, once administered, they had probable cause to arrest Scott and search his living room for weapons.
I respectfully dissent.